Aus dem Funktionsbereich für Parodontologie

(Leiter: Univ.-Prof. Dr. med. dent. Thomas Kocher)

in der Poliklinik für Zahnerhaltung, Parodontologie, Endodontologie

(Direktor: Univ.-Prof. Dr. med. dent. Dr. h.c. Georg Meyer)

im Zentrum für Zahn-, Mund- und Kieferheilkunde

(Geschäftsführender Direktor: Univ.-Prof. Dr. med. dent. Dr. h.c. Georg Meyer)

der Universitätsmedizin der Ernst-Moritz-Arndt-Universität Greifswald

Aspects of oral health in the German Oral Health Studies

Inaugural-Dissertation

zur

Erlangung des akademischen

Grades

Doktor der Wissenschaften in der Medizin

(Dr. rer. med.)

der

Universitätsmedizin

der

Ernst-Moritz-Arndt-Universität

Greifswald

2015

vorgelegt von: Svenja Schützhold

geb. am: 14.01.1986

in: Papenburg

Dekan: Prof. Dr. med. dent. Reiner Biffar

1. Gutachter: Prof. Thomas Kocher

2. Gutachter: Prof. Christoph Bandt

3. Gutachter: PD Dr. Nicole Pischon

Ort, Raum: Bibliothek, Ellernholzstr. 1-2 des Institutes für Community Medicine, Greifswald

Tag der Disputation: 04.12.2015

Table of contents

Page

1 Introduction .................................................................................................................... 1

2 Aims ............................................................................................................................... 3

3 Methods .......................................................................................................................... 5

3.1 Design of included studies .......................................................................................... 5

3.2 Variables and statistical analyses. ............................................................................... 7

4 Results ........................................................................................................................... 9

5 Discussion .................................................................................................................... 14

6 Summary ...................................................................................................................... 18

7 Zusammenfassung ........................................................................................................ 19

8 References .................................................................................................................... 21

Appendix ......................................................................................................................... 24

1

1 Introduction

According to the World Health Organization (WHO), 'oral health is a state of being free from

chronic mouth and facial pain, oral and throat cancer, oral sores, birth defects as cleft lip and

palate, periodontal (gum) disease, tooth decay and tooth loss, and other diseases and disorders

that affect the oral cavity' [1].

The oral health status has a significant impact on daily life [2, 3]. The status of the teeth plays

an important role for smile attractiveness, which, in turn, is associated with self-confidence

and self-esteem [4]. Esthetics, social and professional success were more often attributed to

individuals with normal dentition than to individuals with well-perceivably decayed teeth [5].

The particular importance of the dental status is emphasized by the fact that the presence of a

reduced functional dentition (nine teeth or fewer) was found to be negatively associated with

the physical index of quality of life (QOL) [6]. The presence of nine or fewer teeth in the

maxilla was even ranked on the physical index of QOL between renal diseases and cancer.

Moreover, median Oral Health Impact Profile (OHIP) scores significantly differed between

partially edentulous patients and fully dentate patients [7]. The most frequently reported

problems in the group of partially edentulous patients were chewing difficulties,

psychological disappointment and dissatisfaction with appearance due to problems with teeth,

mouth or dentures. In addition, missing teeth may lead to tooth migration and tipping. The

altered jaw function, as the long-term effect of missing teeth, may be related to the formation

of temporomandibular disorders [8].

Self-perception of oral health by single-item questions is a simple and validated tool to gather

information of people`s perception of oral health [9]. Self-rated oral health captures the

patients` subjective and objective perception of oral health and is strongly associated with

dental care-seeking behaviors [10, 11]. In the past few years, numerous studies included

single-item questions to assess self-perceived oral health [12]. Numbers of missing teeth were

associated with a higher probability of perceiving oral health as poor [12-14]. Subjects who

assessed their oral health as 'poor' had three times as many missing teeth as subjects assessing

their oral health as 'excellent' [15]. It was stated that the variable missing teeth 'is clearly the

most important clinical factor in determining patients` evaluations of their oral health status'

[16]. However, self-perceived general health is not only influenced by the awareness of the

clinical status but also by social and cultural backgrounds, and particularly, by comparisons

with people of a similar age [17].

2

Against this background, a 'natural experiment' such as the German reunification in 1990 is of

great interest to researchers due to the possibility of getting new insights into the determinants

of self-perceived health. In 1992, the proportion of participants reporting less than good

health was significantly lower in East than in West Germany (47.2% versus 53.5%) [18].

Interestingly, the prevalence of poor self-perceived health increased both in East and West

Germany until 1997, with the result that the gap between East and West almost diminished

(56.6% in East and 55.9% in West Germany). The authors ascribe these observations to 'the

view of an 'assimilation' process towards west German attitudes or beliefs in East Germany

after reunification' [18]. After reunification, East Germany adopted the West German health-

care system, which fundamentally differed from its Eastern counterpart. Before reunification,

almost all dentists were salaried employees in East Germany, not receiving any commissions

beyond their salaries. In contrast, in West Germany, there was a system of reimbursement, in

which health insurances paid dentists for each treatment. Great efforts were made after

reunification to bring East Germany into line with the West German social market economy.

Although economic differences diminished over time, East Germans remained

socioeconomically deprived; in 1997 the unemployment rate was 19.1% in East and 10.8% in

West Germany [19]. An international collaborative study examined the impact of different

health-care systems on oral health by comparing data for Hannover, West Germany (1973),

and Leipzig, East Germany (1979) [20]. For 35-44-year-olds, the percentages of missing teeth

were higher in the metropolitan (8.1%) and nonmetropolitan (7.4%) parts of Hannover than in

the metropolitan (4.3%) and nonmetropolitan areas (4.7%) of Leipzig. In comparison with

other European countries, the mean number of missing teeth in Germany ranks in the middle

to high range [21].

A review of the literature on the epidemiology of tooth loss showed a significant decline in

prevalence and incidence of severe tooth loss between 1990 and 2010 at the global, regional,

and country levels [22]. The global age-standardized prevalence of severe tooth loss

decreased from 4.4% (95% CI: 4.1%-4.8%) to 2.4% (95% CI: 2.2%-2.7%), and the incidence

rate decreased from 374 cases (95% CI: 347-406) to 205 cases (95% CI: 187-226) per

100,000 person-years. The higher tooth retention might be explained by comprehensive caries

prevention campaigns during the last decades [23, 24]. Additionally, a higher number of teeth

was associated with lower mean attachment loss (AL) [25]. Periodontitis at least partially

seemed to account for tooth loss [21]. Thus, the positive development towards higher tooth

retention may not only be attributed to the caries decline [26], but also to a possible decrease

of the prevalance and severity of periodontitis. Hence, it is important for health care planners

3

to know whether the prevalence of periodontitis in Germany changed during the last decades.

However, only few studies have examined periodontal trends so far. During the last decades,

most studies referred to an improvement of periodontitis [27-32], whereas some studies

reported improvements of gingivitis and mild/moderate periodontitis without obviously

observable reductions in severe forms of periodontits [33]. Possibly, this pattern represents

today's restraints in extracting periodontally diseased teeth in contrast to previous decades

[34]. For Germany, only few data are currently available, e.g. on the basis of the cross-

sectional and nationally representative German Oral Health Studies ('Deutsche

Mundgesundheitsstudien', DMS). As reported in previous studies, the Community Periodontal

Index (CPI) deteriorated between 1997 and 2005 [35]. However, the CPI has several

shortcomings [36, 37] and does not include AL, which is preferably used to define

periodontitis in combination with probing depth (PD) and/or bleeding on probing [38].

2 Aims

Against this background, the aims of this thesis were as follows:

- Assessment of the relative contributions of clinical oral health variables to self-perceived

oral health by means of an age-specific approach

- Evaluation of the changes in dental health in West and East Germany between 1989 and

2005

- Detailed evaluation of changes in periodontal health and number of teeth within the last

decade based on data from the DMS studies and the Studies of Health in Pomerania (SHIP)

4

This thesis is based on the following three published original articles:

1. Schützhold S, Holtfreter B, Schiffner U, Hoffmann T, Kocher T, Micheelis W (2014).

Clinical factors and self-perceived oral health. European Journal of Oral Sciences 122:134-

141.

2. Schützhold S, Holtfreter B, Hoffmann T, Kocher T, Micheelis W (2013). Trends in dental

health of 35- to 44-year-olds in West and East Germany after reunification. Journal of Public

Health Dentistry 73:65-73.

3. Schützhold S, Kocher T, Biffar R, Hoffmann T, Schmidt CO, Micheelis W, Jordan R,

Holtfreter B (2015). Changes in prevalence of periodontitis in two German population-based

studies. Journal of Clinical Periodontology 42:121-130 (which was awarded the Eugen-

Fröhlich-Preis of the DG PARO 2014).

5

3 Methods

3.1 Design of included studies

The German Oral Health Studies (DMS)

The Institute of German Dentists [Institut der Deutschen Zahnärzte (IDZ)] conducted four

national cross-sectional surveys of oral health in the German resident population (DMS): in

1989 (DMS I, only West Germany), 1992 (DMS II, only East Germany), 1997 (DMS III), and

2005 (DMS IV). In this study, the first two surveys (1989/92) were merged to achieve

comparability with the last two studies [39]. Random cluster samples stratified by Federal

State and by community category were drawn, altogether of 80 (DMS I), 40 (DMS II), 90

(DMS III), and 90 (DMS IV) municipalities. Random samples were selected from the records

of registration offices from each of these municipalities. East Germans were oversampled.

Samples from 35-44-year-olds (adults) were taken in all four DMS studies (DMS I-IV),

whereas samples from 65-74-year-olds (seniors) were only drawn in DMS III and IV. For

adults, participation rates averaged 56%, 72%, 56% and 52%, respectively. Both in DMS III

and IV, the participation rates equaled 56% for seniors (Table 1).

Table 1. Information on sampling design for adults and seniors in DMS I–IV. Survey

name

Birth cohort Survey

year

Age

(year)

Sampling

size

Number of invited

subjects

Participation

rate (%)

Adults

DMS I 1945-1954 1989 35-54* 1700* 1544* 858 (56%)*

DMS II 1948-1957 1992 35-54* 1039* 1014* 731 (72%)*

DMS III 1953-1962 1997 35-44 1260 1179 655 (56%)

DMS IV 1961-1970 2005 35-44 1980 1774 925 (52%)

Seniors

DMS III 1923-1932 1997 65-74 2520 2424 1367 (56%)

DMS IV 1931-1940 2005 65-74 1980 1868 1040 (56%)

* Information was only available for the cohort aged 35-54

6

Studies of Health in Pomerania (SHIP)

The Studies of Health in Pomerania (SHIP) are two independent cross-sectional population-

based studies conducted during 1997-2001 (SHIP-0) and 2008-2012 (SHIP-Trend) in

northeast Germany. In SHIP-0, a two-stage cluster sampling design yielded twelve 5-year-

strata (20-79 years) for both genders, each including 292 subjects. In the first sampling stage,

three cities and 12 larger towns were selected, and then 17 of 97 small villages (

7

3.2 Variables and statistical analyses

Schützhold, S., et al. (2014). "Clinical factors and self-perceived oral health." Eur J Oral

Sci 122(2): 134-41

Self-perceived oral health was evaluated by the question: 'Thinking of your teeth, how would

you rate their condition?' (with the response options: 'very good', 'good', 'satisfactory', 'less

good', or 'poor'). We combined the first and the last two answers and, therefore, the dependent

variable had three categories: very good/good, satisfactory, and less good/poor (subsequently

referred to as good, satisfactory, and poor). We additionally assessed the following clinical

oral health variables: numbers of decayed teeth, numbers of filled teeth, and numbers of

unreplaced teeth, mean AL, bleeding on probing (BOP), the presence of a fixed denture, and

the presence of a removable denture. Separate multinomial logistic regression models were

developed for both age groups in DMS IV (Table 2) to evaluate associations of self-perceived

oral health with clinical oral health variables. Stepwise forward and backward regression

analyses were performed. Following recent recommendations for the conduct of such analyses

[40], we chose p for inclusion as 0.15 and p for exclusion as 0.20. All analyses were

weighted.

Schützhold, S., et al. (2013). "Trends in dental health of 35- to 44-year-olds in West and

East Germany after reunification." J Public Health Dent 73(1): 65-73

Because only 35-44-year-olds provided the longest observation period with examinations

between 1989 (DMS I) and 2005 (DMS IV), analyses were restricted to 35-44-year-old

subjects (Table 2). Numbers of decayed, missing and filled teeth were determined and the

DMFT-index was calculated. The number of sound teeth was calculated as the difference

between 28 and the DMFT-index. Subjects were stratified into those with more versus less

than or equal to 20 teeth. Regression models were applied to assess associations between

region, survey year, their interactions and variables assessing dental disease status (number of

missing, filled, decayed and sound teeth, the DMFT-index and the probability of having ≤20

teeth). Negative binomial regression models were applied to account for the highly positively

skewed counts of decayed and missing teeth. Multiple linear regressions were used to model

the number of filled and sound teeth and the DMFT-index. Logistic models were applied to

model the probability of having ≤20 teeth. To assess effects of region within specific survey

8

years, linear combinations of estimators were tested using Wald tests. Analyses were

weighted and adjusted for potential risk factors for caries, namely gender, age, school

education, marital status, utilization of dental services, number of snacks, use of dental floss /

tooth sticks, and the last dental visit.

Schützhold, S., et al. (2015). "Changes in prevalence of periodontitis in two German

population-based studies." J Clin Periodontol 42(2): 121-30

In SHIP, measurements of AL and PD were taken at distobuccal, midbuccal, mesiobuccal,

and midlingual/palatinal sites according to the half-mouth method excluding third molars

(SHIP-0: alternating on the left or right side; SHIP-Trend: left or right side randomly

selected). In DMS III, AL and PD were determined at midbuccal and mesiobuccal sites in the

first and fourth quadrant. In DMS IV, AL and PD were assessed at midbuccal, mesiobuccal

and distolingual sites at twelve index teeth (17, 16, 11, 24, 26, 27, 47, 46, 44, 31, 36, 37). To

ensure comparability between DMS III and IV, teeth and sites had to be brought down to a

common denominator for further analyses, i.e. at maximum two sites (midbuccal and

mesiobuccal) on six teeth (17, 16, 11, 44, 46, 47). The number of teeth in dentates was

determined excluding third molars.

The prevalence of a given condition, e.g. AL ≥3 mm, was defined as the percentage of

subjects having at least one site with that condition. Extent was defined as the percentage of

teeth displaying that condition. For DMS III and IV, data were presented stratified by

residence (West or East Germany). Because of the complex sample designs of DMS and

SHIP, analyses were weighted. In SHIP-0, design variables were additionally considered to

identify strata and clusters and to adjust for finite population corrections at both sampling

stages. In SHIP-Trend, both sampling weights and the stratification variable were included

into the analyses. To compare distributions of periodontal variables between groups, chi

square tests were applied, whose statistics were corrected for the final sampling weights. To

assess differences in the numbers and percentages of the affected teeth between the studies,

Mann-Whitney U tests were applied, which were performed by use of the Somers' D

parameter to account for the final sampling weights.

The results were considered statistically significant at p

9

4 Results

Schützhold, S., et al. (2014). "Clinical factors and self-perceived oral health." Eur J Oral

Sci 122(2): 134-41

Data of 891 adults (35-44-year-old) and 760 seniors (65-74-year-old) from DMS IV were

evaluated. The proportions of participants reporting poor oral health were 18.5% in adults and

16.1% in seniors. After stepwise regression analysis, the number of unreplaced teeth, the

number of filled teeth, the number of decayed teeth, the presence of a removable denture, and

mean AL were added in this order to the final model for adults (Table 3).

Table 3: Outcome of stepwise multinomial regression analysis for adults (35-44-year-old) in DMS IV

Final multinomial model

Satisfactory vs. Good Poor vs. Good

Variable RRR (95% CI) p value RRR (95% CI) p value

Unreplaced teeth (cont.) 1.27 (1.14; 1.40)

10

Table 4: Outcome of stepwise multinomial regression analysis for seniors (65-74-year-old) in DMS IV

Final multinomial model

Satisfactory vs. Good Poor vs. Good

Variable RRR (95% CI) p value RRR (95% CI) p value

Removable denture (ref. No) 1.00 1.00

Yes 2.18 (1.43; 3.33)

11

Figure 1. Trends in dental health variables for adults (35-44-year-old) in DMS I-IV [(a) number of missing

teeth, (b) number of filled teeth or (c) number of decayed teeth, (d) DMFT-index, (e) predicted probability of

having ≤20 teeth, (f) number of sound teeth] according to region and survey year; * indicates p

12

Schützhold, S., et al. (2015). "Changes in prevalence of periodontitis in two German

population-based studies." J Clin Periodontol 42(2): 121-30

Both in SHIP and DMS, the number of teeth in dentates increased significantly in all age

groups (p

13

In SHIP, the total prevalence of AL ≥3 mm decreased from 89.7% to 85.1% and the

percentage of teeth being affected decreased from 62.8% to 55.9% (p

14

5 Discussion

The evidence generated in this thesis can be summarized as follows:

1. The number of unreplaced teeth showed the strongest association with self-perceived

oral health in adults and was the second variable to enter the model for seniors during the

stepwise selection process.

2. Although the number of missing teeth decreased between 1997 and 2005 in the whole

Republic, East Germany was not able to completely catch up with West Germany.

3. The periodontal status significantly improved in SHIP and in West German adults during

the last decade.

In our study, the number of unreplaced teeth was most strongly associated with self-perceived

oral health. It was the first variable to enter the model for adults and the second variable to

enter the model for seniors. Strong associations between missing teeth and self-perceived oral

health were also observed in other studies [12, 13]. These findings underline the importance

of missing teeth for the self-perception of oral health. Possible reasons are the reduced

chewing efficiency, the unfavorable appearance, and the necessity for prosthetic restorations.

Generally, we proceeded on the assumption that the awareness of adverse oral conditions

translates into poor self-perceived oral health. However, answers to the open-ended question

about factors that influence the self-perception of health revealed that most participants

reported that they compare their health with that of reference groups, for example with 'people

their age' or 'friends/acquaintances' [17]. Social factors, such as culture, class, and race, also

influence the extent of translation of adverse clinical conditions into poor self-perceived oral

health [47].

Tooth loss mainly reflects past oral disease experiences and the access to oral care during

people`s lives. In our study, East Germans had consistently more missing teeth than West

Germans in each survey year. The almost parallel running curves of missing teeth indicate

that even East German participants of DMS IV, having lived 20-30 years in the German

Democratic Republic (GDR) and 15 years in the re-united Federal Republic of Germany,

were not able to catch up with their Western counterparts. The explanation might be found in

15

the completely different health-care systems. Before reunification, almost all East German

dentists were salaried employees in state-owned clinics and availability of material was often

poor [48]. In constrast to the tooth-maintaining attitudes of West German dentists, East

German dentists pursued attitudes toward tooth extractions. These concepts changed after

reunification, as, by the end of 1992, 88% of all East German dentists moved from salaried

employment into private practices [48]. The differences in the reasons for extractions between

East and West Germany evened out just in the last 15 years [49]. Possibly, a complete

convergence between East and West Germany might only be achieved for birth cohorts born

after reunification.

We observed a decrease of the number of missing teeth between 1997 and 2005 in Germany.

A declining trend of tooth loss over the last decades has also been observed in other studies

[27, 28, 30, 50]. Moreover, the authors of a review on edentulism stated that the prevalence of

edentulism will drop by up to 50-60% over the next 20 years in Finland, Sweden and the UK

[51]. The improved tooth retention might be explained by comprehensive caries prevention

campaigns, especially by the increased utilization of fluoridated toothpastes in the last

decades and by the resulting caries decline [26]. Caries is one of the two main reasons for

tooth extractions, the other main factor is periodontitis [49]. In our study, we observed an

improvement of the periodontal status in SHIP and in West German adults during the last

decade. A recently published review article supports the assumption that periodontal disease

prevalence is declining, though to varying degrees [52]. Based on these findings, it can be

expected that the improvement of periodontal status might translate into a even higher tooth

retention in the future.

Besides caries and periodontitis as the two main reasons for tooth extractions, there are other

factors that determine the fate of a tooth, for example dental traumas, orthodontic reasons or

prosthetic treatments [49]. Additionally, social determinants like race and socioeconomic

status can also have strong impacts on tooth loss incidence [53]. Given the same disease

extent and severity, African Americans and persons with lower socioeconomic status were

more likely to receive a dental extraction once they enter the dental care system than non-

Hispanic whites. Additionally, they were less likely to report that dentists had discussed

alternative treatments with them [53]. According to Preshaw et al. [54], 'the patients`

decisions are likely to be influenced by variables such as the strategic location of the tooth,

the importance they place on retaining teeth, their ability (and willingness) to pay for the

16

necessary care that is required of a tooth can be saved, their willingness to undergo treatment,

and the availability of specialist cares to resolve complex issues'. Another important aspect

worth mentioning is the subjectivity of the dentist's clinical decision making. Even when

differences in patients were controlled, the variations in dentists` decisions were omnipresent

[55].

Some strengths and limitations of the studies deserve consideration. A strength of our

analyses is the examination of trends over a time span of 16 years (1989-2005) in DMS,

combined with the concurrent consideration of socioeconomic variables. In addition, the

simultaneous analyses of changes in periodontal status in two German population-based

studies (SHIP and DMS) represents another unique strength of our analyses. One limitation

might be the potential selection bias due to the high percentage of nonresponders in DMS and

SHIP-Trend. Nonresponse analyses in DMS III and IV, which were based on short

questionnaires with basic questions sent to nonresponders, revealed that study participants

were more often women and visited the dentist more frequently than nonresponders [35, 56].

Conversely, in seniors, study participants were more often men. Therefore, periodontal and

carious prevalences might have been affected differentially. However, as there were only

minor differences between responders and nonresponders, the consequences of selection bias

might be negligible. As a consequence, disease prevalences were potentially slightly

underestimated. Another limitation concerns the equalization of the recording protocols of

DMS III and IV in the course of the trend analyses of periodontal prevalences, because, for

that reason, the analyses in DMS were restricted to two sites at six teeth, which led to only a

small number of sites entering the analyses. The medians with percentiles (25%; 75%) for the

number of sites entering the analyses in DMS III were for adults: 10 (8; 12) and for seniors: 6

(4; 8). In DMS IV, the respective numbers were 12 (10; 12) for adults and 8 (4; 10) for

seniors. As opposed to this, analyses in SHIP were more robst because the recording protocols

were identical and did not necessitate any equalization. Another strength of SHIP is the wide

age range (20-84 years). However, the SHIP studies are limited by their regionality and are

therefore not representative for the whole of Germany. Moreover, different periodontal probes

were used, which may have led to a possible overestimation of PD ≥3 mm and an

underestimation of PD ≥4 mm in SHIP-0 as compared to SHIP-Trend [57].

In summary, we observed strong associations between the number of unreplaced teeth and

self-perceived oral health. Overall, the number of unreplaced teeth was the most important

17

factor in multivariate modeling. The awareness of relative contributions of clinical variables

to self-perceived oral health is important for obtaining a clearer understanding of patients`

subjective and objective self-perceptions of oral health. Between 1997 and 2005, the number

of missing teeth considerably decreased in DMS but East Germans had consistently more

missing teeth than West Germans. Further, during the last decade, the periodontal status

improved in SHIP and in West German adults, which might translate into a even higher tooth

retention in the future. The rapidly changing dental and oral health during the last decades

necessitates regular new data to keep up with ongoing changes in the dental community [50].

For that reason the recently completed fifth German Oral Health Study (DMS V) [58] is of

great interest to researchers as it provides valuable data for comparisons with DMS IV. It

remains to be seen how fast the declining trends of tooth loss will develop in the future.

18

6 Summary

The Institute of German Dentists [Institut der Deutschen Zahnärzte (IDZ)] conducted four

national cross-sectional surveys of oral health in the German resident population [German

Oral Health Studies, 'Deutsche Mundgesundheitsstudien', (DMS)]: in 1989 (DMS I, only

West Germany), 1992 (DMS II, only East Germany), 1997 (DMS III), and 2005 (DMS IV).

In this study, the first two surveys (1989/92) were merged to achieve comparability with the

last two studies. The Studies of Health in Pomerania (SHIP) are two independent regional

cross-sectional population-based studies conducted during 1997-2001 (SHIP-0) and 2008-

2012 (SHIP-Trend) in northeast Germany.

In this thesis, we addressed three main questions: First, we aimed to explore the relative

contributions of clinical oral health variables assessing caries, periodontal status, and

prosthetic status to self-perceived oral health by means of an age-specific approach in DMS

IV. Second, we aimed to assess the changes of dental health in West and East Germany

between 1989 and 2005 in DMS I-IV. Third, we aimed to evaluate the changes of periodontal

status and number of teeth within the last decade based on data from the DMS and the SHIP

studies.

To explore the associations of self-perceived oral health with clinical oral health variables, we

developed separate multinomial logistic regression models for adults and seniors in DMS IV

by using stepwise methods. To assess the changes of dental health in West and East Germany

between 1989 and 2005, we applied regression models and assessed associations between

region, survey year, their interactions and variables assessing dental disease status (number of

missing, filled, decayed and sound teeth, the DMFT-index and the probability of having ≤20

teeth), adjusting for potential risk factors for caries. To assess changes of periodontal status in

Germany, prevalences, percentages and numbers of teeth affected were defined.

In summary, the number of unreplaced teeth showed the strongest association with self-

perceived oral health in adults and was the second variable to enter the model for seniors

during the stepwise selection process. Between 1997 and 2005, the number of missing teeth

considerably decreased in DMS but East Germans had consistently more missing teeth than

West Germans in each survey year. Further, during the last decade, the periodontal status

significantly improved in SHIP and in West German adults, which might translate into a even

higher tooth retention in the future.

19

7 Zusammenfassung

Das Institut der Deutschen Zahnärzte (IDZ) führte vier nationale Querschnittstudien über die

Mundgesundheit der deutschen Bevölkerung (Deutsche Mundgesundheitsstudien, DMS) in

1989 (DMS I, nur Westdeutschland), 1992 (DMS II, nur Ostdeutschland), 1997 (DMS III)

und 2005 (DMS IV) durch. Um die Vergleichbarkeit mit den letzten beiden Studien zu

gewährleisten, wurden die ersten beiden Studien (1989/92) zusammengefügt. Die „Studies of

Health in Pomerania“ (SHIP) sind zwei unabhängige, regionale, populationsbasierte

Querschnittstudien, die zwischen 1997 und 2001 (SHIP-0) und zwischen 2008 und 2012

(SHIP-Trend) in Nordostdeutschland durchgeführt wurden.

In dieser Doktorarbeit wurden im Wesentlichen drei Fragestellungen verfolgt: Das erste Ziel

stellte die Untersuchung der Einflüsse klinischer Variablen (Karies, parodontaler und

prothetischer Status) auf die orale Selbstwahrnehmung mithilfe eines altersspezifischen

Ansatzes in der DMS IV dar. Das zweite Ziel war die Einschätzung der Veränderungen der

dentalen Gesundheit sowohl in West- als auch in Ostdeutschland zwischen 1989 und 2005

anhand der DMS I-IV. Unser drittes Ziel bestand in der Bewertung der Veränderungen der

parodontalen Gesundheit und der Zahnzahl innerhalb des letzten Jahrzehnts, basierend auf

den DMS und den SHIP-Studien.

Für die Untersuchung der Assoziationen zwischen klinischen Mundgesundheitsvariablen und

oraler Selbstwahrnehmung wurden, für Erwachsene und Senioren getrennt, multinomiale

logistische Regressionen mit schrittweiser Vorwärts- und Rückwärts-Technik durchgeführt.

Für die Einschätzung der Veränderungen der dentalen Gesundheit in West- und

Ostdeutschland zwischen 1989 und 2005 verwendeten wir Regressionsmodelle (mit

gleichzeitiger Adjustierung für mögliche Risikofaktoren für Karies) und untersuchten die

Assoziationen zwischen der Region, dem Erhebungsjahr, ihren Interaktionen und den

Mundgesundheitsvariablen (Anzahl der fehlenden, der kariösen und der naturgesunden

Zähne, DMFT-Index und die Wahrscheinlichkeit, ≤20 Zähne zu haben). Für die

Untersuchung der Veränderungen des Parodontalstatus in Deutschland wurden Prävalenzen,

relative Ausmaße und absolute Zahlen der betroffenen Zähne bestimmt.

Zusammengefasst konnten wir zeigen, dass die Zahl der fehlenden, nicht ersetzten Zähne die

stärkste Assoziation mit der oralen Selbstwahrnehmung bei den Erwachsenen aufwies. Die

Zahl der fehlenden, nicht ersetzten Zähne stellte zudem die zweite Variable dar, die während

der schrittweisen Regression in das Modell der Senioren aufgenommen wurde. Zwischen

20

1997 und 2005 verringerte sich die Zahl der fehlenden Zähne der Probanden in den DMS

deutlich, allerdings wiesen die Ostdeutschen durchgängig zu allen Erhebungszeitpunkten eine

höhere Anzahl fehlender Zähne auf als die Westdeutschen. Überdies wurde eine signifikante

Verbesserung des Parodontalstatus in SHIP und bei den westdeutschen Erwachsenen in DMS

innerhalb des letzten Jahrzehnts beobachtet, was zu einem noch höheren Zahnerhalt in der

Zukunft führen könnte.

21

8 References

1. WHO. Oral Health. 2014 [cited 2014 25.10.]; Available from:

http://www.who.int/topics/oral_health/en/.

2. Gomes, A.S., C. Abegg, and J.M. Fachel, Relationship between oral clinical conditions and

daily performances. Braz Oral Res, 2009. 23(1): p. 76-81.

3. Dahl, K.E., N.J. Wang, and K. Ohrn, Does oral health matter in people's daily life? Oral health-

related quality of life in adults 35-47 years of age in Norway. Int J Dent Hyg, 2012. 10(1): p.

15-21.

4. Van der Geld, P., et al., Smile attractiveness. Self-perception and influence on personality.

Angle Orthod, 2007. 77(5): p. 759-65.

5. Eli, I., Y. Bar-Tal, and I. Kostovetzki, At first glance: social meanings of dental appearance. J Public Health Dent, 2001. 61(3): p. 150-4.

6. Mack, F., et al., The impact of tooth loss on general health related to quality of life among

elderly Pomeranians: results from the study of health in Pomerania (SHIP-O). Int J

Prosthodont, 2005. 18(5): p. 414-9.

7. Nickenig, H.J., et al., Oral health-related quality of life in partially edentulous patients:

assessments before and after implant therapy. J Craniomaxillofac Surg, 2008. 36(8): p. 477-

80.

8. Tallents, R.H., et al., Prevalence of missing posterior teeth and intraarticular

temporomandibular disorders. J Prosthet Dent, 2002. 87(1): p. 45-50.

9. Thomson, W.M., et al., Construct validity of Locker's global oral health item. J Dent Res, 2012. 91(11): p. 1038-42.

10. Afonso-Souza, G., et al., Association between routine visits for dental checkup and self-

perceived oral health in an adult population in Rio de Janeiro: the Pro-Saude Study.

Community Dent Oral Epidemiol, 2007. 35(5): p. 393-400.

11. Gilbert, G.H., et al., The paradox of dental need in a population-based study of dentate

adults. Med Care, 2003. 41(1): p. 119-34.

12. Locker, D., E.W. Mscn, and A. Jokovic, What do older adults' global self-ratings of oral health

measure? J Public Health Dent, 2005. 65(3): p. 146-52.

13. Wu, B., et al., Differences in self-reported oral health among community-dwelling black,

Hispanic, and white elders. J Aging Health, 2011. 23(2): p. 267-88. 14. Ekanayke, L. and I. Perera, Factors associated with perceived oral health status in older

individuals. Int Dent J, 2005. 55(1): p. 31-7.

15. Matthias, R.E., et al., Factors affecting self-ratings of oral health. J Public Health Dent, 1995.

55(4): p. 197-204.

16. Reisine, S.T. and H.L. Bailit, Clinical oral health status and adult perceptions of oral health.

Soc Sci Med Med Psychol Med Sociol, 1980. 14A(6): p. 597-605.

17. Kaplan, G. and O. Baron-Epel, What lies behind the subjective evaluation of health status?

Soc Sci Med, 2003. 56(8): p. 1669-76.

18. Nolte, E. and M. McKee, Changing health inequalities in east and west Germany since unification. Soc Sci Med, 2004. 58(1): p. 119-36.

19. Bundesagentur für Arbeit. Registered Unemployed, Unemployment rate by area 2011

01.04.2011 [cited 2011 05.12.]; Available from:

https://www.destatis.de/EN/FactsFigures/Indicators/LongTermSeries/LabourMarket/lrarb00

3.html.

20. Arnljot, H.A., et al., Oral health care systems. An international collaborative study. 1985,

Geneva: World Health Organization.

21. Konig, J., B. Holtfreter, and T. Kocher, Periodontal health in Europe: future trends based on

treatment needs and the provision of periodontal services--position paper 1. Eur J Dent Educ,

2010. 14 Suppl 1: p. 4-24.

22

22. Kassebaum, N.J., et al., Global Burden of Severe Tooth Loss: A Systematic Review and Meta-

analysis. J Dent Res, 2014. 93(7 suppl): p. 20S-28S.

23. Holst, D. and A.A. Schuller, Equality in adults' oral health in Norway. Cohort and cross-

sectional results over 33 years. Community Dent Oral Epidemiol, 2011. 39(6): p. 488-97. 24. Skudutyte-Rysstad, R. and H.M. Eriksen, Changes in caries experience among 35-year-old

Oslo citizens, 1973-2003. Acta Odontol Scand, 2007. 65(2): p. 72-7.

25. Holtfreter, B., et al., Epidemiology of periodontal diseases in the Study of Health in

Pomerania. J Clin Periodontol, 2009. 36(2): p. 114-23.

26. Marthaler, T.M., Changes in dental caries 1953-2003. Caries Res, 2004. 38(3): p. 173-81.

27. Kalsbeek, H., et al., Trends in periodontal status and oral hygiene habits in Dutch adults

between 1983 and 1995. Community Dent Oral Epidemiol, 2000. 28(2): p. 112-8.

28. Dye, B.A., et al., Trends in oral health status: United States, 1988-1994 and 1999-2004. Vital

Health Stat 11, 2007(248): p. 1-92.

29. Rothlisberger, B., et al., Periodontal conditions in Swiss army recruits: a comparative study between the years 1985, 1996 and 2006. J Clin Periodontol, 2007. 34(10): p. 860-6.

30. Skudutyte-Rysstad, R., H.M. Eriksen, and B.F. Hansen, Trends in periodontal health among

35-year-olds in Oslo, 1973-2003. J Clin Periodontol, 2007. 34(10): p. 867-72.

31. Hugoson, A., B. Sjodin, and O. Norderyd, Trends over 30 years, 1973-2003, in the prevalence

and severity of periodontal disease. J Clin Periodontol, 2008. 35(5): p. 405-14.

32. Haisman-Welsh, R.J. and W.M. Thomson, Changes in periodontitis prevalence over two

decades in New Zealand: evidence from the 1988 and 2009 national surveys. N Z Dent J,

2012. 108(4): p. 134-8.

33. Gjermo, P.E., Impact of periodontal preventive programmes on the data from epidemiologic studies. J Clin Periodontol, 2005. 32 Suppl 6: p. 294-300.

34. Baelum, V. and R. Lopez, Periodontal disease epidemiology - learned and unlearned?

Periodontol 2000, 2013. 62(1): p. 37-58.

35. Micheelis, W. and U. Schiffner, Vierte Deutsche Mundgesundheitsstudie (DMS IV), ed. Institut

der Deutschen Zahnärzte (IDZ). 2006, Köln: Deutscher Zahnärzte Verlag

36. Baelum, V., et al., Validity of CPITN's assumptions of hierarchical occurrence of periodontal

conditions in a Kenyan population aged 15-65 years. Community Dent Oral Epidemiol, 1993.

21(6): p. 347-53.

37. Gjermo, P., et al., Periodontal diseases in Central and South America. Periodontol 2000, 2002.

29: p. 70-8. 38. Savage, A., et al., A systematic review of definitions of periodontitis and methods that have

been used to identify this disease. J Clin Periodontol, 2009. 36(6): p. 458-67.

39. Micheelis, W. and J. Bauch, Mundgesundheitszustand und -verhalten in Ostdeutschland, ed.

Institut der Deutschen Zahnärzte (IDZ). 1993, Köln: Deutscher Ärzte-Verlag.

40. Hosmer, D.W., Jr., S. Lemeshow, and R.X. Sturdivant, Applied Logistic Regression, Wiley,

Editor. 2013, John Wiley & Sons, Inc.: Hoboken, New Jersey. p. 89-226.

41. StataCorp, Stata Statistical Software: Release 11. 2009, StataCorp LP: College Station, TX.

42. R Development Core Team, R: A language and environment for statistical computing. 2010, R

Foundation for Statistical Computing: Vienna, Austria.

43. R Core Team, R: A Language and Environment for Statistical Computing. 2013, R Foundation for Statistical Computing: Vienna, Austria.

44. Schutzhold, S., et al., Clinical factors and self-perceived oral health. Eur J Oral Sci, 2014.

122(2): p. 134-41.

45. Schutzhold, S., et al., Trends in dental health of 35- to 44-year-olds in West and East Germany

after reunification. J Public Health Dent, 2013. 73(1): p. 65-73.

46. Schutzhold, S., et al., Changes in prevalence of periodontitis in two German population-based

studies. J Clin Periodontol, 2015. 42(2): p. 121-30.

47. Gilbert, L., Social factors and self-assessed oral health in South Africa. Community Dent Oral

Epidemiol, 1994. 22(1): p. 47-51.

23

48. Widstrom, E., et al., Oral healthcare in transition in Eastern Europe. Br Dent J, 2001. 190(11):

p. 580-4.

49. Glockmann, E., et al., "Ursachen des Zahnverlustes in Deutschland"- Dokumentation einer

bundesweiten Erhebung (2007) ["Reasons for tooth loss in Germany"- Documentation of a

nationwide survey (2007)], in IDZ-Information I.o.G.D. (IDZ), Editor. 2011: Köln.

50. Muller, F., M. Naharro, and G.E. Carlsson, What are the prevalence and incidence of tooth

loss in the adult and elderly population in Europe? Clin Oral Implants Res, 2007. 18 Suppl 3: p.

2-14.

51. Mojon, P., J.M. Thomason, and A.W. Walls, The impact of falling rates of edentulism. Int J

Prosthodont, 2004. 17(4): p. 434-40.

52. Holtfreter, B., S. Schützhold, and T. Kocher, Is Periodontitis Prevalence Declining? A Review of

the Current Literature. Current Oral Health Reports, 2014. 1(4): p. 251-261.

53. Gilbert, G.H., R.P. Duncan, and B.J. Shelton, Social determinants of tooth loss. Health Serv

Res, 2003. 38(6 Pt 2): p. 1843-62. 54. Preshaw, P.M., et al., Association of removable partial denture use with oral and systemic

health. J Dent, 2011. 39(11): p. 711-9.

55. Bader, J.D. and D.A. Shugars, Variation in dentists' clinical decisions. J Public Health Dent,

1995. 55(3): p. 181-8.

56. Micheelis, W. and E. Reich, Dritte Deutsche Mundgesundheitsstudie (DMS III), ed. Institut der

Deutschen Zahnärzte (IDZ). 1999, Köln: Deutscher Ärzte-Verlag.

57. Holtfreter, B., et al., Effects of different manual periodontal probes on periodontal

measurements. J Clin Periodontol, 2012. 39(11): p. 1032-41.

58. Jordan, R.A., et al., The Fifth German Oral Health Study (Funfte Deutsche Mundgesundheitsstudie, DMS V) - rationale, design, and methods. BMC Oral Health, 2014.

14(1): p. 161.

24

Appendix

Verzeichnis wissenschaftlicher Arbeiten von Svenja Schützhold

Publikationen

Schützhold S, Kocher T, Biffar R, Hoffmann T, Schmidt CO, Micheelis W, Jordan R, Holtfreter B.

Changes in prevalence of periodontitis in two German population-based studies. Journal of Clinical

Periodontology. 2015. 42(2): 121-130

Holtfreter B, Schützhold S, Kocher T. Is periodontitis prevalence declining? A review of the current

literature. Current Oral Health Reports. 2014. 1(4):251-261

Schützhold S, Holtfreter B, Schiffner U, Hoffmann T, Kocher T, Micheelis W. Clinical factors and

self-perceived oral health. European Journal of Oral Sciences. 2014. 122(2): 134-41.

Graetz C, Schwendicke F, Kahl M, Dörfer C, Sälzer S, Springer C, Schützhold S, Kocher T, König J,

Rühling A. Prosthetic rehabilitation of patients with history of moderate to severe periodontitis: a

long-term evaluation. Journal of Clinical Periodontology. 2013. 40(8):799-806.

Schützhold S, Holtfreter B, Hoffmann T, Kocher T, Micheelis W. Trends in dental health of 35-44-

year-olds in West and East Germany after reunification. Journal of Public Health Dentistry. 2013.

73(1):65-73.

Konferenzbeiträge

Mense M, Schützhold S, Holtfreter B, Kocher T. Verlauf klinischer Parameter und Zahnverlust

während der parodontalen Erhaltungstherapie (Poster). Jahrestagung der Deutschen Gesellschaft für

Parodontologie (DGParo e.V.). 2014. Sep 19. Münster.

Schützhold S, Kocher T, Biffar R, Hoffmann T, Micheelis W, Jordan R, Holtfreter B. Trends in

periodontal disease in two German population-based studies (Vortrag). IADR General Session &

Exhibition and IADR Africa/Middle East Regional Meeting; 2014 Jun26, Kapstadt, Südafrika.

Schützhold S, Mense M, Holtfreter B, Kocher T. Factors associated with tooth loss in supportive

periodontal therapy (Poster). Jahrestagung der Deutschen Gesellschaft für Parodontologie (DGParo

e.V.). 2013. Sep 20. Erfurt.

Schützhold S, Holtfreter B, Hoffmann T, Kocher T, Micheelis W. Trends in oral health in West and

East Germany after reunification (Poster). EuroPerio 7. 2012 Jun 06. Wien. In: Journal of Clinical

Periodontology. 39 (Supplement s13) 1-429.

Kahl M, Schützhold S, Springer C, El-Sayed KM Fawzy, Rühling A, Kocher T, Dörfer C.E., Graetz C.

Influence of the variable age for long-term tooth retention after 15 years of periodontal supportive

therapy (SPT) (Poster). EuroPerio 7. 2012 Jun 06. Wien. In: Journal of Clinical Periodontology. 39

(Supplement s13) 1-429 und im: International Poster Journal (IPJ) 2013, Ausgabe 2.

25

Eidesstattliche Erklärung

Hiermit erkläre ich, dass ich die vorliegende Dissertation selbstständig verfasst und keine

anderen als die angegebenen Hilfsmittel benutzt habe.

Die Dissertation ist bisher keiner anderen Fakultät, keiner anderen wissenschaftlichen

Einrichtung vorgelegt worden.

Ich erkläre, dass ich bisher kein Promotionsverfahren erfolglos beendet habe und dass eine

Aberkennung eines bereits erworbenen Doktorgrades nicht vorliegt.

Datum Unterschrift

26

Danksagung

Allen voran möchte ich Herrn Prof. Thomas Kocher für die freundliche Überlassung des

Themas danken, darüber hinaus für die intensive Betreuung meiner Doktorarbeit, für die

stetige Unterstützung und für das entgegengebrachte Vertrauen. Außerdem möchte ich mich

für die zahlreichen Möglichkeiten, Kongresse und Tagungen zu besuchen, bedanken.

Einen ganz besonderen Dank möchte ich Dr. Birte Holtfreter aussprechen. Birte, danke für

deine unerschöpfliche Geduld und für dein immer offenes Ohr! Ebenso möchte ich meiner

Bürokollegin Christiane für die angenehme Arbeitsatmosphäre und für die vielen schönen

Momente in unserem Büro danken!

Ich danke Dr. Christian Schwahn für die immer bereitwillige Bereitstellung seiner Literatur,

für die anregenden Gespräche und für die kritischen Denkanstöße.

Ich danke dem Institut der Deutschen Zahnärzte (IDZ) für das Forschungsstipendium, das es

mir ermöglichte, diese Doktorarbeit zu schreiben. Ganz herzlich danke ich auch den

Mitwirkenden und Teilnehmern der Study of Health in Pomerania (SHIP) und der Deutschen

Mundgesundheitsstudien (DMS), die dadurch das Fundament für meine Doktorarbeit schufen.

Ich danke von Herzen meinen Eltern, die mir das Studium, das die Grundlage für diese

Promotion darstellte, ermöglicht haben. Ihnen und auch meinen Freunden danke ich für die zu

jeder Tageszeit erfolgten Aufmunterungen.

Meinem Freund Thomas danke ich für sein unglaubliches Maß an Rückhalt. Ohne seine

liebevolle Unterstützung und ohne seine unendlich großen Motivationshilfen besonders in der

Endphase der Dissertation hätte sich die Durchführung dieser Arbeit weitaus schwieriger

gestaltet.

27

Anlage der Originalarbeiten zur Dissertationsschrift

1. Schützhold, S., et al. (2014). "Clinical factors and self-perceived oral health." Eur J Oral

Sci 122 (2): 134-41

2. Schützhold, S., et al. (2013). "Trends in dental health of 35- to 44-year-olds in West and

East Germany after reunification." J Public Health Dent 73(1): 65-73

3. Schützhold, S., et al (2015). "Changes in prevalence of periodontitis in two German

population-based studies." J Clin Periodontol 42(2): 121-30

Clinical factors and self-perceived oral

health

Sch€utzhold S, Holtfreter B, Schiffner U, Hoffmann T, Kocher T, Micheelis W.

Clinical factors and self-perceived oral health.

Eur J Oral Sci 2014; 122: 134–141. © 2014 Eur J Oral Sci

Self-perceived oral health is affected not only by awareness of the clinical status but

also by comparisons with people of a similar age. This study explored the relative

contributions of clinical variables assessing caries, periodontal status, and prosthetic

status to self-perceived oral health within two age groups. Data of 891 adults (35–

44 yr of age) and 760 older people (65–74 yr of age) from the Fourth German Oral

Health Study (DMS IV, 2005) were evaluated. Self-perceived oral health was

obtained from questionnaires. Numbers of decayed, filled, and unreplaced teeth,

mean attachment loss, bleeding on probing (BOP), the presence of a fixed denture,

and the presence of a removable denture were assessed. Multinomial logistic regres-

sion models were developed for both age groups, separately, using stepwise meth-

ods. For adults, unreplaced teeth, filled teeth, decayed teeth, the presence of a

removable denture, and mean attachment loss were added to the final model. For

older people, the presence of a removable denture, unreplaced teeth, decayed teeth,

mean attachment loss, filled teeth, and BOP were included in the final model.

Awareness of the relative contributions of clinical variables to self-perceived oral

health is important for obtaining a clearer understanding of patients’ subjective and

objective self-perceptions of oral health.

Svenja Sch€utzhold1, BirteHoltfreter1, Ulrich Schiffner2,Thomas Hoffmann3, ThomasKocher1, Wolfgang Micheelis4

1Unit of Periodontology, Department ofRestorative Dentistry, Periodontology andEndodontology, University Medicine,Ernst-Moritz-Arndt-University Greifswald,Greifswald, Germany; 2Department ofRestorative and Preventive Dentistry, Schoolof Dental and Oral Medicine, Hamburg,Germany; 3Department of Periodontology,Medical Faculty Carl Gustav Carus, Dresden,Germany; 4Institute of German Dentists, K€oln,Germany

Svenja Sch€utzhold, Unit of Periodontology,Department of Restorative Dentistry,Periodontology and Endodontology, UniversityMedicine, Ernst-Moritz-Arndt-UniversityGreifswald, Rotgerberstr. 8, 17475Greifswald, Germany

E-mail: svenja.schuetzhold@uni-greifswald.de

Key words: adult; caries; German oral healthstudy; older people; self-perceived oral health

Accepted for publication December 2013

Recently, single-item questions of oral health have beenincreasingly included in dental surveys (1). As self-

perceived oral health reflects patients’ subjective andobjective assessments of their oral health, it is highlyassociated with the patient’s perception of treatmentneeds and thus with the demand for dental services.

Those who regularly visit a dentist for routine dentalcheck-ups are more likely to assess their oral health asgood (2). Conversely, people who report poor oral health

have a less marked dental care-seeking behavior (3).In recent years, numerous studies have assessed self-

perceived oral health by single-item questions, and a

number of clinical factors were found to be associatedwith self-perceived oral health (1, 4–8). With respect tocaries status, some studies reported that the numbers of

missing and decayed teeth were proportional to theprobability of rating oral health as poor (1, 8, 9).Conversely, larger numbers of filled teeth were associ-ated with a higher probability of perceiving oral health

as good (1, 9). Across the single-item responses, caries,tooth loss, and periodontitis experience showed mainlyconsistent gradients (8).

The understanding of periodontitis as a silent diseasewas questioned recently (10) when significant associationsbetween periodontal disease severity and self-

perceived oral health were found. Results consistent withthis finding were also reported in other studies (1, 4).Although the sensitivity of single self-reported items in

diagnosing periodontal disorders is low (11), combina-

tions of demographic measures and self-report oral healthquestions perform well in predicting periodontitis (12).

Self-perceptions of general health depend on socialand cultural backgrounds and are affected by prevailingsocial and medical ideologies (13). Thus, self-perceptionof health varies among social groups (13) and is partly

based on comparisons with others, especially with peo-ple of a similar age (14). Different social backgroundsof age groups inherently influence people’s understand-

ing of ‘normality’ (15).Only a few studies have identified the most predictive

clinical factors for self-perceived oral health using an

age-specific approach. Therefore, the aim of this studywas to explore the relative contributions of clinicalvariables assessing caries, periodontal status, and pros-

thetic status to self-perceived oral health among partici-pants 35–44 and 65–74 yr of age from representativesamples of the German resident population.

Material and methods

Study design

The national cross-sectional Fourth German Oral HealthStudy (‘Vierte Deutsche Mundgesundheitsstudie’, DMSIV) was conducted by the Institute of German Dentists(IDZ) in 2005. Random cluster samples stratified byFederal State and by community category were drawnfrom 90 municipalities. Random samples were selected

Eur J Oral Sci 2014; 122: 134–141DOI: 10.1111/eos.12117Printed in Singapore. All rights reserved

� 2014 Eur J Oral Sci

European Journal ofOral Sciences

from the population registry of each of these municipalities.People from East Germany were oversampled. Informedconsent was obtained from all participants. The partici-pation rates for adults (35–44 yr of age) and for olderpeople (65–74 yr of age) were 52.1% and 55.7%, respec-tively. The low response rates might be explained by thefact that examiners were present at each sample point foronly 2 days. Therefore, it was difficult to arrange conve-nient appointments for study participants. Design,sampling, and non-response analyses have previously beendescribed in detail (16–19).

Dependent variable

Self-perceived oral health was assessed by the followingquestion: ‘Thinking of your teeth, how would you ratetheir condition?’ (with the response options: ‘very good’,‘good’, ‘satisfactory’, ‘not so good’, or ‘poor’). After evalu-ation of frequency distributions, the first two answers werecombined into ‘very good/good’ and the last two answerswere combined into ‘not so good/poor’. Therefore, thedependent variable had three categories: very good/good,satisfactory, and not so good/poor (referred to subse-quently as good, satisfactory, and poor).

Covariates

Sociodemographic and behavioral variables, including age,gender (female/male), region (West Germany/East Ger-many), school education (10 yr), smoking status(never/former/current), cohabiting (no/yes), and dental visitwithin the last 12 months (yes/no), were obtained fromquestionnaires. In addition, self-perceived general healthwas assessed by the question: ‘How would you rate yourgeneral health status?’ with the possible answers (in thisorder): ‘very good’, ‘good’, ‘satisfactory’, ‘not so good’, or‘poor’. The multimorbidity score was defined as the sum ofall positive answers to questions assessing self-reportedgeneral systemic diseases, namely hypertension, circulatorydisorders of the heart, myocardial infarction, cardiac insuffi-ciency, stroke, varicose veins, bronchial asthma, gastritis,gastric or duodenal ulcer, diabetes mellitus with or withoutinsulin treatment, cancer, arthrosis, inflammatory joint orspine disease, osteoporosis, mental illness, or the presence ofother diseases not mentioned in the questionnaire. The mul-timorbidity score was considered as a continuous variableand ranged from 0 to 17. A participant was given a multi-morbidity score of 0 if none of the above questions wasanswered with ‘yes’. Conversely, the highest score, of 17,was achieved if all the questions were answered with ‘yes’.All interviews were conducted as verbal and personal face-to-face interviews by specially trained interviewers.

Oral status

Decayed and filled teeth were assessed visually in afull-mouth examination (20). The number of missing teeththat had not been replaced with either a fixed/removabledenture or by an implant were determined. These teeth aresubsequently referred to as unreplaced teeth. Teeth wereextracted because of caries and other reasons. Thirdmolars were excluded from the analysis. Because the distri-bution of the number of decayed teeth was skewed, partici-pants were classified into those with no decayed teethversus those with at least one decayed tooth. Filled andunreplaced teeth were considered as continuous variables.

Periodontal status was assessed by measuring attach-ment loss and calculation of the papillary bleeding index(PBI) (21). Periodontal measurements were performed at 12index teeth (17, 16, 11, 24, 26, 27, 47, 46, 44, 31, 36, and 37;two-digit notation according to the FDI World Dentalfederation), using a periodontal probe (PCP 11.5 WHOprobe; M1W Dental, B€udingen, Germany). Attachment losswas measured at three sites per tooth (mesiobuccal, midbuc-cal, and distolingual). The PBI was assessed at two sites pertooth (buccal and lingual). To avoid collinearity problems,probing pocket depth was not considered.

Mean attachment loss was calculated and tertiles weredetermined because linearity assumptions in regressionmodels were not met. Tertiles were specific for both agegroups (adults: 1st tertile, �0.91 to 2.28; 2nd tertile, 2.29–2.97; and 3rd tertile, 2.99–8.81; and older people: 1sttertile, 1.08–3.50; 2nd tertile, 3.53–4.67; and 3rd tertile,4.70–11.67). The PBI was dichotomized as present orabsent bleeding at site level, and the percentage of siteswith bleeding on probing (BOP) was determined. Bleedingon probing was considered as a continuous variable.

To evaluate the prosthetic status, participants were clas-sified into those with no fixed denture versus those with atleast one fixed denture and into those with no removabledenture versus those with at least one removable denture.

Study participants

Of 925 adults originally included in the study, there wasno information on self-perceived oral health for four, whowere therefore excluded. Additionally, we excluded eightedentulous participants and three participants withoutperiodontal measurements. Nineteen other participantshad missing covariate data, leaving 891 adult participantsfor analyses. Of 1,040 older people originally included inthe study, there was no information on self-perceived oralhealth for 29, who were therefore excluded. Moreover, weexcluded 215 edentulous participants and four participantswithout periodontal measurements. Thirty-two other par-ticipants had missing covariate data, leaving 760 older par-ticipants for analyses.

Statistical analyses

Chi-square tests and Kruskal–Wallis tests were applied toanalyze differences in various variables among participantswith good, satisfactory, and poor self-perceived oralhealth. Multinomial logistic regression was applied toassess associations of self-perceived oral health with clini-cal oral health variables. Multinomial logistic regression isan extension of the binary logistic regression for outcomevariables with more than two categories. A separate binarylogit is estimated for each pair of outcome categories (22).Good self-perceived oral health was chosen as the refe-rence category.

Separate regression models were developed for each agegroup. Stepwise forward and backward regression analyseswere performed considering all clinical oral health vari-ables. To test for all categories of one categorical variablesimultaneously, overall Wald tests were used. P for inclu-sion was 0.15 and P for exclusion was 0.20 (23). Theanalyses were weighted. For each step, areas under recei-ver-operating characteristic (ROC) curves (AUC) weredetermined. Because the dependent variable had threecategories, the following AUCs (with 95% CI) werecalculated: AUC separating good from satisfactory/poor,

Self-perception of oral health 135

AUC separating satisfactory from good/poor, and AUCseparating good/satisfactory from poor. To internally vali-date final models, bootstrap analyses with 200 replicationswere performed and optimism-corrected estimates of AUCvalues were calculated.

The results were considered statistically significant atP ≤ 0.05. Statistical analyses were performed with STATA/SE11.0 (24).

Results

The proportions of participants reporting poor oralhealth were 18.5% in adults and 16.1% in older people

(Table 1). Sociodemographic and behavioral variableswere not significantly associated with self-perceived oralhealth, except for dental visit within the last 12 monthsin older people (P < 0.001). In contrast, the distribu-

tion of self-perceived oral health differed significantlyacross categories of all clinical oral health variables inboth age groups (Table 2).

Stepwise regression models

In adults, all clinical oral health variables, except forthe presence of a fixed denture and BOP, were includedin the final model after stepwise regression analysis

(Table 3). The variables entered the model in the fol-lowing order: unreplaced teeth, filled teeth, decayedteeth, the presence of a removable denture, and meanattachment loss. The AUC values quantifying the sepa-

ration between good and satisfactory/poor for stepwise

multinomial regression models increased from 0.61(the first step of stepwise regression analysis) to 0.71(the final model; Table 3). The AUC values quantifyingthe separation between satisfactory and good/poor

increased from 0.55 to 0.60, and the AUC values quan-tifying the separation between good/satisfactory andpoor increased from 0.60 to 0.70. After bootstrap

analyses, values for optimism were 0.01, 0.02, and 0.02,respectively. Thus, optimism-corrected estimates ofAUC values for the final model were 0.70, 0.58, and

0.68, respectively.In older people, all clinical oral health variables were

included in the final model, except for the presence of afixed denture (Table 4). The variables entered the

model in the following order: the presence of a remo-vable denture, unreplaced teeth, decayed teeth, meanattachment loss, filled teeth, and BOP. The AUC values

increased from 0.58 to 0.68 for the separation betweengood and satisfactory/poor and from 0.51 to 0.61 forthe separation between satisfactory and good/poor

(Table 4). The AUC values for the separation betweengood/satisfactory and poor increased from 0.62 to 0.72.After bootstrap analyses, values for optimism were

0.02, 0.03, and 0.02, respectively. Thus, optimism-corrected estimates of AUC values for the final modelwere 0.66, 0.58, and 0.70, respectively.

Sensitivity analyses in older people

After addition of the multimorbidity score to the final

model (Table 5), similar coefficients for clinical oral

Table 1

Distribution of socio-economic and behavioral variables according to self-perceived oral health and age group

Variable

Self-perceived oral health of adults (35–44 yr of age)

Self-perceived oral health of older people

(65–74 yr of age)

n Good Satisfactory Poor P n Good Satisfactory Poor P

n 891 40.9 40.6 18.5 760 38.4 45.5 16.1

Age 891 39.5 � 2.7 39.8 � 3.0 39.4 � 2.7 0.29 760 68.8 � 2.6 68.6 � 2.7 68.8 � 2.8 0.50

Gender

Female 503 42.4 37.8 19.8 392 35.7 44.9 19.4

Male 388 38.6 43.0 18.4 0.34 368 40.5 45.1 14.4 0.17

Region

West Germany 591 39.7 40.0 20.3 508 37.4 44.5 18.1

East Germany 300 44.3 42.2 13.5 0.055 252 40.2 47.1 12.7 0.18

School education

10 yr 292 44.3 38.6 17.1 0.47 156 45.1 37.2 17.7 0.20

Smoking status

Never 400 44.3 39.5 16.2 474 39.4 45.1 15.5

Former 178 38.9 44.0 17.1 237 35.4 45.9 18.7

Current 313 36.7 39.4 23.9 0.10 49 36.6 40.8 22.6 0.65

Cohabiting

Non-cohabiting 171 37.5 43.8 18.7 158 38.3 41.1 20.6

Cohabiting 720 41.2 39.6 19.2 0.64 602 37.9 46.1 16.0 0.37

Dental visit within last 12 months

Yes 809 41.6 40.4 18.0 690 39.2 45.9 14.9

No 82 30.7 40.5 28.8 0.06 70 26.6 37.6 35.8

health variables were observed. The multimorbidityscore was significantly associated with self-perceivedoral health (P < 0.001) when poor self-perceived oral

health was compared with good self-perceived oralhealth. The corresponding AUC values increased by0.01 for the separation between satisfactory and good/poor and by 0.02 for the separation between good/sat-

isfactory and poor.After addition of self-perceived general health to the

final model (Table 6), similar coefficients for clinical

oral health variables were observed. Self-perceivedgeneral health was significantly associated with self-perceived oral health in all categories (P < 0.05). For

each separation, the respective AUC values increasedby 0.06.

Discussion

This study explored the relative contributions of clinical

characteristics (such as caries, periodontal status, andprosthetic status) to self-perceived oral health amongadult (35–44 yr of age) and older (65–74 yr of age)

participants in the DMS IV study. In multivariate mode-ling, the number of unreplaced teeth showed the strong-est association with self-perceived oral health in adults,

Table 2

Distribution of clinical oral health variables according to self-perceived oral health and age group

Variable

Self-perceived oral health of adults (35–44 yr of age)

Self-perceived oral health of older people

(65–74 yr of age)

n Good Satisfactory Poor P n Good Satisfactory Poor P

n 891 40.9 40.6 18.5 760 38.4 45.5 16.1

Decayed teeth

0 675 44.8 40.2 15.0 588 41.3 43.4 15.3

>0 216 27.1 41.1 31.8

whereas in older people the presence of at least oneremovable denture was the most important factor.

Although periodontal variables were included in thefinal model in both age groups, they were not as stronglyassociated with self-perceived oral health as were vari-

ables assessing caries and prosthetic status. The AUCvalues showed an acceptable ability of classification(23), indicating an acceptable perception of clinical oral

health variables by both adults and older people.Some limitations deserve consideration. First, self-

perceived oral health was assessed using a single-item

question. Although multi-item scales are more

consistent and stable (25), single-item questions haveadvantages over multi-item scales in that they are cost-

effective and easily interpretable (25). Furthermore, asingle-item question, similar to the one used in thisstudy, was successfully validated in a recent study of 35-

to 44-yr-old adults (8). Here, caries, tooth loss, and peri-odontitis experiences showed mostly consistent gradi-ents across single-item responses. Second, analyses were

restricted to dentates for whom periodontal measureswere available. Consequently, the numbers of adultsand older people were reduced by 3.7% and 26.9%,

respectively. Especially in older people, the association

Table 4

Outcome of stepwise multinomial regression analysis for older people (65–74 yr of age)

Variable

Final multinomial model Discriminatory power for stepwise built models

Satisfactory versus Good Poor versus Good AUCs (95% CI) for the three classes

RRR (95% CI) P RRR (95% CI) P Good Satisfactory Poor

Removable denture

No (ref.) 1.00 1.00

Yes 2.18 (1.43–3.33)

between the presence of a removable denture and self-perceived oral health might have been underestimated.Third, the periodontal recording protocol required

assessment of attachment loss only at three sites pertooth, at 12 index teeth. Compared with a full-mouthprotocol, index teeth include a high percentage of

molars and mandibular incisors, which are more oftenafflicted by periodontitis (26). Thus, mean attachmentloss might have been slightly overestimated in the pres-ent study. The bias caused by assessing only three sites

per tooth, instead of six sites, might be negligible.Fourth, we used mean attachment loss to assess peri-odontal disease experience. Although mean attachment

loss estimates are robust against probe- or examiner-associated bias, their use might reduce much of the in-terindividual variation. Extent measures of attachment

loss would not reduce much of the interindividual varia-tion. However, because both estimates performed com-paratively well regarding their discriminative power for

the specific outcome, we decided to use mean attach-ment loss in this study. Finally, the high percentage ofnonresponders might have led to selection bias. Shortbasic questionnaires sent to nonresponders revealed that

adult nonresponders were more often men (54.7% vs.50.6%) and visited the dentist for regular check-ups lessfrequently (64.9% vs. 76.1%). Conversely, in older peo-

ple, nonresponders were more often women (57.9% vs.53.8%). Thus, periodontal and carious prevalencesmight have been affected differentially. However, as

there were only minor differences between respondersand nonresponders, selection bias might be marginal.

We generally assume that awareness of adverse oral

conditions implies poor self-perceived oral health.

However, answers to an open-ended question aboutfactors influencing the self-perception of health revealedthat most participants reported comparing their health

with that of reference groups, especially with that oftheir age peers (13). Social factors, such as culture,class, and race, also influence the extent of translation

of adverse clinical conditions into poor self-perceivedoral health (15).

Many participants were satisfied with their oralhealth; only 18.5% of adults and 16.1% of older people

reported having poor oral health. Apparently, olderpeople consider an impaired oral health status asbeing part of the ordinary aging process (27, 28). In

our study, sensitivity analyses revealed that the multi-morbidity score was significantly associated with self-perceived oral health in older people when poor

self-perceived oral health was compared with good self-perceived oral health. Apparently, the presence of othermedical diseases was also associated with self-perceived

oral health in older people. The more medical diseasesreported, the higher the likelihood that participantsrated their oral health as poor.

Overall, the number of unreplaced teeth was the

most important factor. It was the first variable to enterthe model for adults and the second variable to enterthe model for older people. Strong associations between

missing teeth and self-perceived oral health were alsoobserved in other studies (1, 9, 29). Possible reasons arethe reduced chewing efficiency, the unfavorable appear-

ance, and the necessity for prosthetic restorations.Factors assessing caries status were highly predictive.

In both age groups, the number of decayed teeth was

the third variable to enter the model. This strong

Table 6

Outcome of sensitivity analyses (including self-perceived general health) after stepwise multinomial regression analysis for olderpeople (65–74 yr of age; n = 758a)

Variable

Final multinomial model Discriminatory power for stepwise built models

Satisfactory versus Good Poor versus Good AUCs (95% CI) for the three classes

RRR (95% CI) P RRR (95% CI) P Good Satisfactory Poor

Removable denture

No (ref.) 1.00 1.00

Yes 2.06 (1.33–3.20) 0.001 5.02 (2.79–9.04)

association is consistent with other studies (1, 9). Inadults, both bivariate and multinomial analyses revealedthat the probability of assessing oral health as poorincreased proportionally to the number of filled teeth. In

contrast, older people with a higher number of filledteeth tended to rate their oral health as good or satisfac-tory; however, this effect was severely diminished in mul-

tivariate regression. Discrepancies between these agegroups might be explained by the fact that older peoplewith removable dentures have fewer natural teeth,

which, in turn, implies a low maximum number ofpossibly filled teeth. Additionally, filled teeth alsoencompassed crowned teeth, which are needed to attachremovable dentures, and thus they were perceived as

being necessary and inevitable. Other studies consistentlyreported that in older participants a higher number offilled teeth correlated with a higher probability of per-

ceiving oral health as good (1, 9). These findings oncemore underline the age-dependent shift of perception.

Mean attachment loss was included in the final model

in both age groups but it was not as strongly associatedwith self-perceived oral health as caries and prostheticfactors. The AUC values only slightly improved after the

addition of mean attachment loss in both age groups. Inthe literature, periodontal disease is often considered as asilent disease (11, 30). In contrast, other studies foundthat periodontal inflammation (6) and mean attach-

ment loss (1, 4) were significantly associated with self-perceived oral health. Moreover, models combiningscreening questions for periodontitis and risk factors

performed well in predicting periodontitis (12, 31). All inall, periodontal problems might indeed be perceived byparticipants, but to a much lesser extent than caries and

prosthetic factors.In summary, the number of unreplaced teeth was

most strongly associated with self-perceived oral health

in adults, whereas for older people the presence of aremovable denture was the most important factor. Inboth age groups, periodontal status was not as stronglyassociated with self-perceived oral health as were caries

and prosthetic status, which indicates that periodontaldisease severity does not contribute to self-perceivedoral health in multivariate models. Awareness of the

relative contributions of clinical variables to self-perceived oral health is important for gaining a clearerunderstanding of the patients’ subjective and objective

self-perceptions of oral health.

Acknowledgements – DMS IV was funded by the two Federal den-tal organizations, the Bundeszahn€arztekammer/BZ€AK and theKassenzahn€arztliche Bundesvereinigung/Cologne. S.S. was sup-ported by the Institut der Deutschen Zahn€arzte/IDZ (Institute ofGerman Dentists). B.H. was financed by an unlimited educationalgrant from GABA, Switzerland.

Conflicts of interest – There are no conflicts of interest associatedwith this work.

References

1. LOCKER D, WEXLER EW, JOKOVIC A. What do older adults’global self-ratings of oral health measure? J Public HealthDent 2005; 65: 146–152.

2. AFONSO-SOUZA G, NADANOVSKY P, CHOR D, FAERSTEIN E,WERNECK GL, LOPES CS. Association between routine visitsfor dental checkup and self-perceived oral health in an adultpopulation in Rio de Janeiro: the Pro-Saude Study. CommunDent Oral Epidemiol 2007; 35: 393–400.

3. GILBERT GH, SHELTON BJ, CHAVERS LS, BRADFORD EH Jr.The paradox of dental need in a population-based study ofdentate adults. Med Care 2003; 41: 119–134.

4. GIFT HC, ATCHISON KA, DRURY TF. Perceptions of the natu-ral dentition in the context of multiple variables. J Dent Res1998; 77: 1529–1538.

5. ANDRADE FB, LEBRAO ML, SANTOS JL, DUARTE YA, TEIXEIRADS. Factors related to poor self-perceived oral health amongcommunity-dwelling elderly individuals in Sao Paulo, Brazil.Cad Saude Publica 2012; 28: 1965–1975.

6. NORTHRIDGE ME, CHAKRABORTY B, KUNZEL C, METCALF S,MARSHALL S, LAMSTER IB. What contributes to self-rated oralhealth among community-dwelling older adults? Findingsfrom the ElderSmile program. J Public Health Dent 2012; 72:235–245.

7. PATTUSSI MP, PERES KG, BOING AF, PERES MA, DA COSTAJS. Self-rated oral health and associated factors in Brazilianelders. Commun Dent Oral Epidemiol 2010; 38: 348–359.

8. THOMSON WM, MEJIA GC, BROADBENT JM, POULTON R. Con-struct validity of Locker’s global oral health item. J Dent Res2012; 91: 1038–1042.

9. WU B, PLASSMAN BL, LIANG J, REMLE RC, BAI L, CROUT RJ.Differences in self-reported oral health among community-dwelling black, Hispanic, and white elders. J Aging Health2011; 23: 267–288.

10. CUNHA-CRUZ J, HUJOEL PP, KRESSIN NR. Oral health-relatedquality of life of periodontal patients. J Periodontal Res 2007;42: 169–176.

11. DIETRICH T, STOSCH U, DIETRICH D, SCHAMBERGER D, BERN-IMOULIN JP, JOSHIPURA K. The accuracy of individual self-reported items to determine periodontal disease history. Eur JOral Sci 2005; 113: 135–140.

12. EKE PI, DYE BA, WEI L, SLADE GD, THORNTON-EVANS GO,BECK JD, TAYLOR GW, BORGNAKKE WS, PAGE RC, GENCORJ. Self-reported measures for surveillance of periodontitis. JDent Res 2013; 92: 1041–1047.

13. KAPLAN G, BARON-EPEL O. What lies behind the subjectiveevaluation of health status? Soc Sci Med 2003; 56: 1669–1676.

14. KRAUSE NM, JAY GM. What do global self-rated health itemsmeasure? Med Care 1994; 32: 930–942.

15. GILBERT L. Social factors and self-assessed oral health inSouth Africa. Commun Dent Oral Epidemiol 1994; 22: 47–51.

16. HOLTFRETER B, KOCHER T, HOFFMANN T, DESVARIEUX M,MICHEELIS W. Prevalence of periodontal disease and treat-ment demands based on a German dental survey (DMS IV).J Clin Periodontol 2010; 37: 211–219.

17. SCHIFFNER U, HOFFMANN T, KERSCHBAUM T, MICHEELIS W.Oral health in German children, adolescents, adults andsenior citizens in 2005. Community Dent Health 2009; 26:18–22.

18. GEYER S, SCHNELLER T, MICHEELIS W. Social gradients andcumulative effects of income and education on dental healthin the fourth German oral health study. Commun Dent OralEpidemiol 2010; 38: 120–128.

19. MICHEELIS W, SCHIFFNER U. Vierte Deutsche Mundgesundhe-itsstudie (DMS IV). K€oln: Deutscher Zahn€arzte Verlag,2006.

20. WORLD HEALTH ORGANIZATION. Oral health surveys. Basicmethods, 4th edn. Geneva: World Health Organization, 1997.

21. SAXER UP. Muhlemann HR [Motivation and education]. SSOSchweiz Monatsschr Zahnheilkd 1975; 85: 905–919.

22. LONG JS, FREESE J. Regression Models for Categorical Depen-dent Variables Using Stata, , 2nd edn. College Station, Texas:Stata Press, 2006.

23. HOSMER JR DW, LEMESHOW S, STURDIVANT RX. AppliedLogistic Regression. In: Wiley, ed. Hoboken. New Jersey:John Wiley & Sons, 2013; 89–226.

24. STATACORP. Stata Statistical Software: Release 11. CollegeStation, TX: StataCorp LP, 2009.

140 Sch€utzhold et al.

25. BOWLING A. Just one question: If one question works, why askseveral? J Epidemiol Community Health 2005; 59: 342–345.

26. ALBANDAR JM, BRUNELLE JA, KINGMAN A. Destructive peri-odontal disease in adults 30 years of age and older in theUnited States, 1988-1994. J Periodontol 1999; 70: 13–29.

27. MACENTEE MI, HOLE R, STOLAR E. The significance of themouth in old age. Soc Sci Med 1997; 45: 1449–1458.

28. SLADE GD, SANDERS AE. The paradox of better subjectiveoral health in older age. J Dent Res 2011; 90: 1279–1285.

29. REISINE ST, BAILIT HL. Clinical oral health status and adultperceptions of oral health. Soc Sci Med Med Psychol MedSociol 1980; 14A: 597–605.

30. GILBERT AD, NUTTALL NM. Self-reporting of periodontalhealth status. Br Dent J 1999; 186: 241–244.

31. SLADE GD. Interim analysis of validity of periodontitisscreening questions in the Australian population. J Periodon-tol 2007; 78: 1463–1470.

Self-perception of oral health 141

Trends in dental health of 35- to 44-year-olds in West andEast Germany after reunificationSvenja Schützhold, MSc1; Birte Holtfreter, PhD1; Thomas Hoffmann, PhD2; Thomas Kocher, PhD1;Wolfgang Micheelis, PhD3

1 Unit of Periodontology, University Medicine, Ernst-Moritz-Arndt University Greifswald, Greifswald, Germany

2 Department of Periodontology, Medical Faculty Carl Gustav Carus, Dresden, Germany

3 Institute of German Dentists, Köln, Germany

Keywords

national oral health studies; dental caries;

adult; epidemiologic studies; cross-sectional

survey.

Correspondence

Ms. Svenja Schützhold, Unit of Periodontology,

University Medicine, Ernst-Moritz-Arndt

University Greifswald, Rotgerberstr. 8, 17475

Greifswald, Germany. Tel.: +49-3834-867397;

Fax: +49-3834-867171; e-mail:

svenja.schuetzhold@uni-greifswald.de. Svenja

Schützhold, Birte Holtfreter, and Thomas

Kocher are with the Unit of Periodontology,

University Medicine, Ernst-Moritz-Arndt

University Greifswald. Thomas Hoffmann is

with the Department of Periodontology,

Medical Faculty Carl Gustav Carus. Wolfgang

Micheelis is with the Institute of German

Dentists.

Received: 5/10/2012; accepted: 11/20/2012.

doi: 10.1111/jphd.12007

Journal of Public Health Dentistry 73 (2013) 65–73

Abstract

Objectives: The German reunification (1990) resulted in huge social upheavals inEast Germany involving changes in health-care systems. We aimed to assess thechanges of dental health between 1989 and 2005, hypothesizing that dental healthconverged in West and East Germany.Methods: We evaluated data from 855 East and 1,456 West Germans aged 35-44years from the cross-sectional German Oral Health Studies (Deutsche Mundgesund-heitsstudien) conducted in 1989/92, 1997, and 2005. Regression models were appliedto assess associations between region, survey year, their interactions and variablesassessing dental disease status [number of decayed (DT), missing (MT), and filledteeth (FT), the DMFT-index, the probability of having �20 teeth and the number ofsound teeth (ST)], adjusting for potential risk factors for caries.Results: After a slight increase of MT between 1989/92 and 1997 (West: 3.6 to 3.6;East: 4.5 to 4.9), numbers of MT considerably decreased between 1997 and 2005(West: 3.6 to 2.2; East: 4.9 to 3.1). East Germans had consistently more MT.Numbers of FT, DT, ST, and the DMFT-index equalized at the latest in 2005. TheEast German DMFT-index increased between 1989/92 and 1997 and slightlydecreased between 1997 and 2005, whereas the West German DMFT-index steadilydecreased between 1989/92 and 2005.Conclusions: Dental health converged in West and East Germany, but the highernumber of MT in 2005 indicates that East Germany was not able to catch up com-pletely with West Germany.

Introduction

Before the reunification in 1990, both the former FederalRepublic of Germany (FRG) (old Federal States, referred to asWest Germany) and the former German Democratic Repub-lic (GDR) (new Federal States, referred to as East Germany)had completely different political, economic, and health-caresystems. Until now, these inequalities decreased due to hugeefforts made after reunification to bring East Germany intoline with the West German social market economy. Todevelop a modern infrastructure and to promote a competi-tive economy, the rebuilding of East Germany was financedwith high amounts of transfer payments. This led to anincrease of the East German gross domestic product to 71percent of the West German average in 2008 (1). Although

economic differences diminished over time, East Germansremained socioeconomically deprived; in 1997 the unem-ployment rate was 19.1 percent in East and 10.8 percent i